Meteorology and Atmospheric Physics

, Volume 119, Issue 3, pp 151–161

Disparity in the characteristic of thunderstorms and associated lightning activities over dissimilar terrains

Authors

    • Department of Atmospheric SciencesUniversity of Calcutta
  • Anirban Middey
    • Department of Atmospheric SciencesUniversity of Calcutta
Original Paper

DOI: 10.1007/s00703-012-0226-4

Cite this article as:
Chaudhuri, S. & Middey, A. Meteorol Atmos Phys (2013) 119: 151. doi:10.1007/s00703-012-0226-4

Abstract

Thunderstorms and associated lightning flash activities are studied over two different locations in India with different terrain features. Lightning imaging sensor (LIS) data from 1998 to 2008 are analyzed during the pre-monsoon months (March, April and May). The eastern sector is designated as Sector A that represents a 2° × 2° square area enclosing Kolkata (22.65°N, 88.45°E) at the centre and covering Gangetic West Bengal, parts of Bihar and Orissa whereas the north-eastern sector designated as Sector B that also represents a 2° × 2° square area encircling Guwahati (26.10°N, 91.58°E) at the centre and covering Assam and foot hills of Himalaya of India. The stations Kolkata and Guwahati are selected for the present study from Sector A and Sector B, respectively, as these are the only stations over the selected areas having Radiosonde observatory. The result of the present study reveals that the characteristics of thunderstorms over the two locations are remarkably different. Lightning frequency is observed to be higher in Sector B than Sector A. The result further reveals that though the lightning frequency is less in Sector A, but the associated radiance is higher in Sector A than Sector B. It is also observed that the radiance increases linearly with convective available potential energy (CAPE) and their high correlation reveals that the lightning intensity can be estimated through the CAPE values. The sensitivity of lightning activity to CAPE is higher at the elevated station Guwahati (elevation 54 m) than Kolkata (elevation 6 m). Moderate resolution imaging spectrometer (MODIS) data products are used to obtain aerosol optical depth and cloud top temperature and employed to find their responses on lightning radiance.

Abbreviations

AOD

Aerosol optical depth

CAPE

Convective available potential energy

Cb

Cumulonimbus

CTT

Cloud top temperature

FPM

Flash per minute

LIS

Lightning imaging sensor

MODIS

Moderate resolution imaging spectrometer

TRMM

Tropical rainfall measuring mission

1 Introduction

The concern about natural disasters and severe weather like thunderstorm activities and lightning flashes necessitates the extensive research in this field. Over the Indian sub-continent, eastern and north-eastern parts are more prone to thunderstorms than the other parts of the country (IMD 1941; Guha and De 2009). Though thunderstorms may occur throughout the year (Manohar et al. 1999; Tyagi 2007), the thunderstorms of pre-monsoon period are most severe among all (Chaudhuri and Aich Bhowmick 2006). Every year immense damages on human lives, property, aviation and agriculture are observed due to high wind speed, lightning flashes, occasional hail and tornadoes and pulse rain associated with thunderstorms which cause enormous socio-economic impacts over the region. Initially, Koteswaram and Srinivasan (1958) pioneered the thunderstorm research over Gangetic West Bengal and concluded that concurrent existence of low level convergence and upper level divergence is the key factor for thunderstorm genesis. Mukherjee et al. (1964) studied extensively the thunderstorms over Guwahati and adjoining parts of north-eastern India and observed maximum frequency of thunderstorms during pre-monsoon months (March–May) particularly at night times. Kumar and Mohapatra (2006) observed that average of 5.4 squalls prevails during the pre-monsoon season over the region. Majority of thunderstorms was observed to occur during midnight to early morning over Guwahati and during afternoon to early night over Kolkata. The significance of storm updraft and strength can be well indicated by lightning flash rate (Baker et al. 1995). Reeve and Toumi (1999) observed that a change in the average land wet-bulb temperature of just 1 K over the globe would result in a change in lightning activity of about 40 %. There are number of climate studies using lightning and global electric circuit as global change indicator (Williams 1992; Price 1993; Gatlin and Goodman 2010). Williams (1985) did a pioneering work and established the dependence of lightning activity on the vertical development of convection. Gatlin and Goodman (2010) suggested lightning jump algorithm as a useful operational diagnostic tool for severe thunderstorm potential. The severe and most intense thunderstorm cells among the interacting storm cells within a mesoscale convective system (MCS) can be well diagnosed with lightning activity (Steiger et al. 2007). The microstructure of the cloud affects the lightning activity and the structure of the convective system (Timothy and Rutledge 2002; Khain et al. 2008). Various studies in this field have been done till date (Jacobson et al. 2007; Pessi et al. 2009; McCaul Jr. et al. 2009; Chaudhuri 2008a, b, 2010a, b, 2011a, b; Chaudhuri and Middey 2009, 2011a, b). Positive correlation between convective rainfall and lightning activity over central India has been identified during pre-monsoon months (Lal and Pawar 2009). The variations of lightning activity during tropical monsoon over three islands were studied by Ramesh Kumar and Kamra (2010) and concluded that the strengthening of updraft causes the increase in flash rate in a storm. Significant link between sea surface temperature in the bordering seas (Arabian sea and Bay of Bengal) and lightning activity over peninsular India has been established (Tinmaker et al. 2010) and on a seasonal time scale, two maxima of lightning activity one in May and another in September are found.

In the present study, an attempt is made to identify the disparity in the characteristics of thunderstorms and the associated lightning activities during the pre-monsoon season over the two different locations in India with different terrain features. Lightning imaging sensor (LIS) database from 1998 to 2008 is analyzed for the purpose. The eastern sector, Sector A (21.41°–24.55° latitude, 85.94°–89.3° longitude) covers the Gangetic West Bengal, adjoining part of Bihar and Orissa and the north-eastern sector, while Sector B (24.41°–26.6° latitude, 90.08°–94.04° longitude) covers Assam, foot hills of Himalaya. The aerosol optical depth (AOD) at 550 nm and cloud top temperature (CTT) are taken from moderate resolution imaging spectro-radiometer (MODIS) terra satellite to explore the microphysical effects on lightning activity and convective systems (Blyth et al. 2001). The present study can be used to categorize thunderclouds on the basis of microphysical characteristics, because it is evident that with lower AOD in thunderclouds of Kolkata the radiance of the lightning increases while the reverse situation prevails over Guwahati. The dissimilar microphysical characteristics during thunderstorm activity associated with lightning over Kolkata and Guwahati are the salient observations of the present study.

2 Materials and methods

The lightning data have been collected from the lightning imaging sensor (LIS) database during the period from 1998 to 2008 for the pre-monsoon months (March, April and May). The LIS is a TRMM satellite-based instrument and detects cloud-to-ground (CG) and intra cloud lightning activities within the troposphere. The TRMM satellite has circled the earth at an altitude of 350 km above the ground with an inclination of 35° since 1998. The TRMM satellite was boosted from an altitude of 350–402 km in August 2001 to extend its mission life. The primary component of LIS is a 128 × 128 charged coupled device (CCD) matrix, which has a sampling rate slightly greater than 500 frames per second. This lightning sensor locates and detects lightning flashes with storm-scale resolution (4–7 km) over a large region (600 km × 600 km) of the earth’s surface. The LIS sensor requires approximately 49 days to return to its original position relative to the Sun and the Earth, though the manual recommends using a 100-day window (Boccippio et al. 1998). LIS has a much smaller field of view, but has higher detection efficiency. A real time event processor (RTEP) removes the background signal and determines the time of occurrence, even in the presence of bright sunlit clouds, thus enabling the system to detect weak lightning and achieve 90 % detection efficiency (Koshak et al. 2000). The flash optical radiance observed by LIS is a measure of lightning discharge energy or the discharge current. In the present study, the data from the moderate resolution imaging spectro-radiometer (MODIS; King et al. 1992) are used to study the cloud top temperature (CTT) and aerosol optical depth (AOD) at 550 nm channel. The data set for March, April and May from 1998 to 2008 is collected from version 5 of MODIS terra satellite over Sector A and Sector B. The pre-monsoon thunderstorm records are collected from India Meteorological Department (IMD).

3 Results and discussions

The characteristics of thunderstorms associated with lightning activities over the two sectors are observed to be completely different. Sector A covers major parts of Gangetic West Bengal, part of Bihar plateau (Jharkhand) and portion of Orissa (Fig. 1). Sector A mainly comprises plain lands over the bank of the river Ganges and its tributaries. It has been observed that the combined effect of Bay of Bengal in the south-east and Chotanagpur plateau in the north-west parts of the sector plays a key role in the initiation of thunderstorm over Kolkata during the pre-monsoon season (Mukhopadhyay et al. 2009). The changes in meteorological parameters, before and after the occurrence of thunderstorms, were described using statistical frequency domain analyses (Chaudhuri and Biswas 2009). Various studies on thunderstorms over Assam and adjoining region during April and May 2006 were made under a pilot project entitled ‘severe thunderstorm observation and regional modeling’ (STORM) and revealed that maximum frequency of thunderstorm is along Brahmaputra river (Chakrabarti et al. 2008). The cardinal component for the occurrence of such thunderstorms is observed to be the lower level cyclonic circulation over sub-Himalayan West Bengal and adjoining west Assam which moves towards east under the influence of trough in westerly and finally strike Guwahati and other regions of Assam. Sector B has a very different terrain features than that of Sector A (Fig  1). The lower and central Assam include Khasi, Garo and Jaintia hill ranging from Shillong plateau which has heights between 3,000 and 6,000 ft. The alluvial plains of Assam and foothills of Himalaya consist of the valley of the Brahmaputra and its tributaries. The station Guwahati is situated within the southern bank of the Brahmaputra river and the foothills of the Shillong plateau. The database from 1998 to 2008 is analyzed during pre-monsoon months (March, April and May) over the two sectors (Fig. 1). The average numbers of flashes during the pre-monsoon months over Sector A show maxima in the month of May (Fig. 2), while in Sector B the maximum number of flashes is observed in the month of April (Fig. 3). This difference may be due to the fact that maximum number of thunderstorms occurs over Sector A during the month of May whereas for Sector B the maximum number of thunderstorms is reported in month of April. However, the significant observation is that the minimum value of lightning flashes over Sector B (~1,500) is almost double than that observed in Sector A (~600). Thus, Sector B is observed to be more lightning prone zone than Sector A. This is quite apparent because the number of thunderstorm events over Sector A (on an average 15) is much less than that of Sector B (on an average 28) during the pre-monsoon season. Records of thunderstorms collected from the Regional Meteorological Center at Kolkata and Guwahati airport clearly affirm the higher frequency of thunderstorm occurrence over Sector B than that of Sector A (Fig. 4). The basic dynamics for the genesis of thunderstorms over the two sectors are observed to be quite different. In Sector A, convective mechanism favours the lifting process for cloud formation, while in Sector B the orographic lifting is the prime triggering mechanism. The flash per minute (FPM) is found to be maximum in the month of April for both the sectors (Figs. 5, 6), but Sector B represents higher value (0.72 FPM) than Sector A (0.51 FPM). The decadal trends of total number of flashes over the sectors reveal significant difference. It is observed that the total flashes increase rapidly since 1998 during the months of April and May, while in the month of March the increase is quite less in Sector A (Fig. 7). On the contrary, March and April show an increasing trend, while the month of May shows a decreasing trend since 1998 in Sector B (Fig. 8). The trends in flash counts can be explained by estimating the frequency of thunderstorms over the two Sectors (A and B). The trend between lightning flash counts and thunderstorm frequencies over the Sector A (Figs. 7, 9) and Sector B (Figs. 8, 10) depicts similar patterns. More number of flash counts is, thus, due to more number of thunderstorm activity and vice versa. The increase or decrease in the total flash counts can, therefore, be explained by estimating the thunderstorms frequency. Convective available potential energy (CAPE), on the other hand, is considered as the measure of thunderstorm potential (Chaudhuri 2010a). The CAPE values are computed using the upper air Radiosonde (RS)/Rawinsonde (RW) sounding data. The increasing trends in CAPE and the flash rate is found to be linearly correlated over both the stations Kolkata (22.65°N, 88.45°E) and Guwahati (26.10°N, 91.58°E) within Sectors A and B, respectively (Fig. 11). It is observed from the figure that with the same value of CAPE, the value of FRM is higher at Guwahati than Kolkata. The average flash optical radiance during the pre-monsoon months over Sectors A and B is compared (Fig. 12). Lightning radiances are found to be much higher in the months of March and April over the Sector A, however, a reverse trend is observed in May whereas Sector B depicts a higher value than Sector A. The month of May is in the transition of the end of pre-monsoon season and the onset of summer monsoon, thus the severe, towering, cumulonimbus cloud activity is supposed to get reduced and mid-level monsoon cloud takes over in Sector A. The flash optical radiance, which is a measure of lightning discharge energy, thus, gets reduced during the month of May over Sector A. However, in Sector B, thunderstorm activity is not affected in the month of May. The lightning radiance increases linearly with CAPE value and high correlation coefficient reveals that lightning intensity can be estimated with CAPE values. The scatter plots of CAPE and lightning radiance over the two stations Kolkata and Guwahati depict a significant linear association between CAPE and lightning radiance with correlation coefficient of 0.82 and 0.85 for Kolkata (22.65°N, 88.45°E) and Gauhati (26.10°N, 91.58°E), respectively (Figs. 13, 14). The linear correlation between CAPE and the lightning radiance may be due to the higher updraft velocity of the parcel with higher available energy. The higher the value of CAPE, higher will be the updraft velocity in the thundercloud. Such increase in updraft speed can bring more particulate matters inside the cloud which favours more charge destruction. As a consequence lightning radiance increases. The linear regression equations relating lightning radiance and CAPE are expressed as:
https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig1_HTML.gif
Fig. 1

The pictures showing the locations of the selected Sector A enclosing the station Kolkata and Sector B enclosing the station Guwahati of India

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig2_HTML.gif
Fig. 2

The frequency of average flash rate of lightning activity over the eastern Sector A enclosing the station Kolkata, India during the pre-monsoon season

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig3_HTML.gif
Fig. 3

The average flash rate of lightning activity over the north-eastern Sector B enclosing the station Guwahati, India during the pre-monsoon

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig4_HTML.gif
Fig. 4

Diagram showing the frequency of thunderstorm occurence over Kolkata and Guwahati for the pre-monsoon months during 1998–2008

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig5_HTML.gif
Fig. 5

The average flash rate per minute over the eastern Sector A enclosing the station Kolkata, India during the pre-monsoon season

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig6_HTML.gif
Fig. 6

The average flash rate per minute over the north-eastern Sector B enclosing the station Guwahati, India during pre-monsoon season

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig7_HTML.gif
Fig. 7

The diagram showing the 10 years trend of total number of flashes during the pre-monsoon months over Sector A enclosing the station Kolkata, India

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig8_HTML.gif
Fig. 8

The diagram showing the 10 years trend of total number of flashes during the pre-monsoon months over Sector B enclosing the station Guwahati, India

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig9_HTML.gif
Fig. 9

Diagram showing the trend in thunderstorm occurrences in the pre-monsoon months during 1998–2008 over Sector A

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig10_HTML.gif
Fig. 10

Diagram showing the trend in thunderstorm occurrences in the pre-monsoon months during 1998–2008 over Sector B

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig11_HTML.gif
Fig. 11

The variation of flash rate per minute with CAPE over two selected stations Kolkata and Guwahati of Sectors A and B, respectively

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig12_HTML.gif
Fig. 12

The lightning radiance over two Sectors (A and B) during the pre-monsoon thunderstorm activity

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig13_HTML.gif
Fig. 13

The scatter plot showing the variation of CAPE with lightning radiance over the station Kolkata of Sector A during pre-monsoon season

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig14_HTML.gif
Fig. 14

The scatter plot showing the variation of CAPE with lightning radiance over the station Guwahati of Sector B during pre-monsoon season

For the station Kolkata,
$$ Y = 0.1083X + 244.74 $$
(1)
For the station Guwahati,
$$ Y = 0.0264X + 190.67, $$
(2)
where Y is the lightning radiance (predictand) and X is CAPE (predictor). The coefficient of determination (R2) of regression equations (1) and (2) is 0.688 and 0.733, respectively (Figs. 13, 14).
A contour plot shows the variation of lightning radiance with cloud top temperature and aerosol optical depth during thunderstorm days over Kolkata (Fig. 15). It is observed that at low cloud top temperature (<210 K) and low aerosol optical depth (<0.5), the intensity of lightning radiance becomes higher at Kolkata (Sector A). The contour plot with same parameters is plotted for Guwahati (Sector B) during pre-monsoon thunderstorm days (Fig. 16). It is found that high cloud top temperature (>260 K) and high aerosol optical depth (>0.7) favour the high intensity of lightning radiance over Guwahati. Low cloud top temperature represents high, towering cumulonimbus (Cb) cloud formation during thunderstorm activity over Kolkata whereas the station Guwahati, being situated at foothills of Assam valley, frequently generates thunderstorms with low cloud heights (as evident from high cloud top temperature). Lightning radiances are representative of charge destroyed per lightning flash. The highest observed lightning radiance over Guwahati (700 kJ/m2/sr) is less than that observed over Kolkata (1,000 kJ/m2/sr). The variation in lightning radiance with cloud top temperature and CAPE during thunderstorm days over Kolkata (22.65°N, 88.45°E) shows high intensity of lightning radiance with low cloud top temperature (<210 K) and high CAPE value (>5,000 J/kg) (Fig. 17). On the contrary, a broad range of cloud top temperature (230–270 K) and low CAPE (<3,000 J/kg) is observed to favour the high intensity of lightning radiance over Guwahati (Fig. 18). It is observed that for thunderstorms in the vicinity of Kolkata lightning frequency is less than that of Guwahati. However, the radiance associated with lightning is more for these thunderstorms than that around Guwahati. The main reason for this is that the thunderstorms in and around Kolkata are convectively generated and have low cloud top temperature and low aerosol optical depth. These categories of thunderclouds are, thus, towering cumulonimbus clouds with higher density of particles within it. The thunderstorms in and around Guwahati are generated due to orographic lifting with higher cloud top temperature and higher optical depth. The number of thunderstorms is less in and around Kolkata as compared to Guwahati.
https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig15_HTML.gif
Fig. 15

The contour plot showing the variation of lightning radiance with cloud top temperature and aerosol optical depth during thunderstorm days over Kolkata (22.65°N, 88.45°E)

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig16_HTML.gif
Fig. 16

The contour plot showing the variation of lightning radiance with cloud top temperature and aerosol optical depth during thunderstorm days over Guwahati (26.10°N, 91.58°E)

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig17_HTML.gif
Fig. 17

The contour plot showing the variation of lightning radiance with cloud top temperature and CAPE during thunderstorm days over Kolkata (22.65°N, 88.45°E)

https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig18_HTML.gif
Fig. 18

The contour plot showing the variation of lightning radiance with cloud top temperature and CAPE during thunderstorm days over Guwahati (26.10°N, 91.58°E)

Two severe thunderstorms on 21 May 2007 and 20 April 2007 are observed to pass over Sectors A and B, respectively. Total granules with flashes detected by LIS in the area of interests are shown in Fig. 19. It can be summarized from the entire analyses that the convective thunderstorms dominate over Sector A as evident by high range of CAPE values and low cloud top temperature. The triggering mechanism is mainly surface low pressure system and upper level divergence. However, over Sector B, the triggering mechanism is due to the orographic lifting. Sector B is in the north-eastern part of India covered by mountains and valley. The thunderstorm occurs in this region with much warmer cloud, having average cloud top temperature around 255 K much higher than that observed in Sector A (average cloud top temperature ~230 K).
https://static-content.springer.com/image/art%3A10.1007%2Fs00703-012-0226-4/MediaObjects/703_2012_226_Fig19_HTML.gif
Fig. 19

Total flashes detected by LIS in the area of interest during thunderstorm events on 21 May 2007 over Sector A and 20 April 2007 over Sector B

4 Conclusion

The lightning activity (using LIS database) during the pre-monsoon months over two geographically different locations in India is studied with different thermodynamic and microphysical parameters; CAPE, cloud top temperature and aerosol optical depth collected from MODIS terra products. The genesis condition and dynamics for the development of thunderstorms over the two locations are observed to be completely different. Lightning activity is higher in Sector B than Sector A. Though the lightning frequency is comparatively less but the associated radiance is higher in Sector A than Sector B. Lightning radiances are representative of charge destroyed per lightning flash, and over Sector A these charge destruction processes are higher than that in Sector B. This may be due to the excess particulate matter present in the atmosphere over Sector A and during entrainment process the particulate matters get injected inside the clouds and because of radiative cooling the charge destruction processes facilitates (Kar et al. 2009). This is also evident from Figs. 15 and 16 that only at higher values of AOD the lightning radiances are higher over Sector B, but high lightning radiances are observed over Sector A even at low AOD. Special reference to the stations Kolkata (22.65°N, 88.45°E) and Guwahati (26.10°N, 91.58°E) belong to Sectors A and B, respectively, is made for thunderstorm and lightning activities. It is, thus, suggested that to forecast the lightning intensity over Sectors A and B, different schemes should be adopted for two zones as it is evident from the present study that lightning intensity and lightning radiance respond to the thermodynamic and dynamic parameters differently within Sectors A and B.

Acknowledgments

The first author acknowledges the India Meteorological Department (IMD), Govt. of India for providing the thunderstorm records and the respective websites for making the meteorological and satellite data available for scientific endeavour. The authors thank the anonymous reviewers for favourable comments and constructive suggestions which helped to improve the clarity of the paper.

Copyright information

© Springer-Verlag Wien 2012